Ai-Driven Currency Forecasting in Emerging African Economies: Risks of Algorithmic Dependence

Artificial Intelligence (AI) has become one of the defining technologies of the 21st century. Despite growing attention to AI applications in currency forecasting, there remains a scarcity of research on African contexts. Specifically, the risks of algorithmic dependence have been underexplored in the African financial policy literature, even as reliance on foreign-developed technologies accelerates. Therefore, this study sought to examine the risks of algorithmic dependence in AI-driven currency forecasting in emerging African nations. The study was grounded in dependency theory. In addition, the study relied on a convergent mixed-methods design. The study gathered secondary data and primary data from 60 key informant interviews (KIIs) with central bankers, fintech practitioners, analysts, and academics from four African nations (Kenya, Egypt, Nigeria, and South Africa). Four risks emerged as primary concerns: limited accountability (22/60; 36.7%), data bias and misrepresentation (16/60; 26.7%), erosion of institutional capacity (13/60; 21.7%), and geopolitical dependence (9/60; 15.0%). The findings revealed that algorithmic dependence is both a technical vulnerability and a socio-political phenomenon that reproduces external epistemic authority and geo-economic asymmetries. Therefore, this study concludes that the promise of AI-driven currency forecasting could turn into a new cycle of subordination in the global financial order if a context-sensitive AI adoption strategy is not implemented for currency forecasting in African states.